期刊名称：The Prague Bulletin of Mathematical Linguistics
印刷版ISSN：0032-6585
电子版ISSN：1804-0462
出版年度：2011
卷号：95
期号：1
页码：33-50
DOI：10.2478/v10108-011-0003-4
语种：English
出版社：Walter de Gruyter GmbH
摘要：In this paper, we present machine learning approach for the classification indirect anaphora in Hindi corpus. The direct anaphora is able to find the noun phrase antecedent within a sentence or across few sentences. On the other hand indirect anaphora does not have explicit referent in the discourse. We suggest looking for certain patterns following the indirect anaphor and marking demonstrative pronoun as directly or indirectly anaphoric accordingly. Our focus of study is pronouns without noun phrase antecedent. We analyzed 177 news items having 1334 sentences, 780 demonstrative pronouns of which 97 (12.44 %) were indirectly anaphoric. The experiment with machine learning approaches for the classification of these pronouns based on the semantic cue provided by the collocation patterns following the pronoun is also carried out.
Loading...
